ZANCTA

/tools/image

Image OCR — extract text locally

Extract text from JPG, PNG, and WebP images in your browser. English OCR is free. Local OCR Power adds Hindi, Bengali, Tamil, Spanish, French, German, and scanned PDF OCR (up to 20 pages) for Premium. Files and recognized text stay on your device.
Local processingPremium paths

Tool details

local boundary
Formats
JPG, PNG, WEBP, PDF
Limits
Up to 50 MB per file · 1 file
Processing
Processing stays in this browser after the page loads.
Result
Download or copy the result from this tab. Nothing is stored on ZANCTA.

Local OCR workspace

OCR is not human-level. Results may vary with image quality, fonts, and layout. Handwriting, tables, and low-contrast scans often fail.

Local processing

Processed in this tab. File bytes are not uploaded. How local processing works

Language

Free OCR — English images. Current image limits apply.

Premium OCR — Hindi, Bengali, Tamil, Spanish, French, German, and scanned PDF OCR (up to 20 pages). Language packs load only when selected. Recognition is not human-level.

Select an image to begin.

Scanned PDF OCR runs in your browser. English OCR is included with the site, and additional language data loads only when selected by a Premium customer. File bytes and extracted text are not sent to an OCR service.

Free, no upload — pass it on:

Before you start

A clear path through Image OCR.

Review the input, the visible limits, and the expected result before you choose a file. The workspace remains the primary place to do the task.

  • Check the input

    JPG, PNG, WEBP, PDF supported · up to 50 MB per file · up to 1 file.

  • Keep the boundary visible

    Selected file bytes are processed in this browser for this supported local workflow.

  • Review the result

    Wait until the tool finishes, inspect the output, then download or clear the workspace.

Related reading: How browser OCR works without uploading

FAQ

Do images leave my device?
No. The OCR engine runs in a browser Web Worker. Selected file bytes and recognized text are not sent to an OCR API.
Which languages are available?
English is free and bundled. Hindi, Bengali, Tamil, Spanish, French, and German load only for Premium accounts, and only when selected.
Can it OCR a scanned PDF?
Yes, as Local OCR Power for Premium, limited to 20 pages. Text-native PDFs should use PDF Text Extractor instead.
How does browser OCR work?
The engine and language data run in this tab. The browser OCR guide explains the privacy boundary and the current limits.

Processing runs locally in your browser — no upload. Outputs are generated on-device. For the privacy boundary see the local processing guide.